Context
By November 2018, 188 individuals fulfilled inclusion and exclusion criteria, agreed to participate, and were randomized. Response rate of eligible and invited individuals was 49%. Most named reasons for declining participation were already active; time constraints; long distance to health care center; have no interest or because of health reasons. Fig 2. provides the number of participants invited, declined, eligible and consented and randomized at the respective primary care center.
Table 1 describes baseline characteristics of the randomized participants per allocated group.
For the total sample, most were diagnosed with type 2 diabetes and body mass index (BMI) levels classified 46% of the participants as obese (≥30 kg/m2). Fewer than 1/3 were classified as having a sedentary lifestyle with an average of 6566 steps per day. About half had a university education and less than 10% were current smokers. No significant differences between the groups were observed at baseline. Characteristics of participants from each primary care center are provided in Appendix to illustrate the characteristics of participants from the different contexts.
Implementation
Fidelity, delivery, and reach
The intervention components were delivered as planned with minor adjustments. Physical activity on prescription was applied occasionally as the diabetes specialist nurses felt it did not add any value above the use of self-monitoring of steps in combination with motivational interviewing. The ratings and logs of motivational interviewing were missing for more than half of the sample and evaluation of the quality of the MI sessions was not possible. Two of the diabetes specialist nurses experienced the consultations as difficult to rate, but all three were in overall satisfied with the consultations. The group sessions followed the planned program, except that one session was cancelled for one group during summer. Some issues during the intervention were reported: low attendance at group sessions (due to work, illness or other obligations); the individual consultations occasionally discussed health matters rather than support for physical activity; technical problems with the pedometers; lost passwords; and temporary problems accessing the website. The technical problems were solved instantly, and these issues were not considered affecting the delivery of the intervention, but to be regarded as natural deviations in clinical practice. No adaptations were made to the study protocol during the intervention, although the participants could use another step counter such as a Fitbit or smart phone if they felt the pedometer was not reliable.
The flow chart in Fig 3 shows dose received (participants’ adherence to the intervention components) by respective primary care center. Mean percentage of days with registered steps on the website over 12 months was 88% for both group A and B. In group A 73% and in group B 65% had an individual step goal. No adverse health events due to participation were reported during the first 12 months intervention.
Physical activity behavior
Mean differences in steps from baseline to 6- and 12 months for each allocated group are listed in table 2. Daily steps increased to a similar extent between baseline and 6 months for Groups A and B. Between baseline and 12 months group A did not change and Group B increased mean daily steps marginally. Steps declined for Group C at each measurement period. Changes in minutes of accelerometer wear-time at each measurement period was negligible for all groups.
Figure 4 and 5 display the percentage of participants in each allocated group reaching 5000 respectively 7000 steps per day at baseline, 6- and 12 months.
At least 60% of participants in all groups maintained at least 5000 steps per day and nearly 40% reached 7000 steps per day at the baseline, 6- and 12-month assessment periods. Groups A and B showed increases in number of participants reaching the 5000 and 7000 steps per day thresholds at the 6-months however these increases declined at 12 months. The proportion of Group C taking at least 5000 steps per day was consistent across the intervention period but declined at 12 months for 7000 steps per day.
The distribution of step change for respective allocated intervention groups are shown in intervals of 1000 steps from baseline to 6 months in figure 6 and from baseline to 12 months in figure 7. There was a spread in the level of step change among the participants in each group and the distribution of changes in the 1000 steps appeared normally distributed for each group at both assessment periods. At 6 months, most participants increased from 1 to 999 steps per day, with an exception of Group B where 10 participants increased by 3000 steps per day or more. At 12 months, a decreased by 0 to 999 steps per day was the most common change.